Tunemeet

Summary

Web app where you can chat with other people who are listening to the same song right now in Spotify.

Tunemeet automatically connects you to other users in the site who are listening to the same song right now. Once matched, a private chat will be created, where you can chat with the other person.

You can also select whether you want the matches to be permament, or temporary. With permament matches the history and connection will be saved and you can come back to it whenever you want. With temporary matching the chat stays active until one of the parties decides to leave it.

Background / Comments

In 2020, Spotify launched a micro site called "Listening Together", where they visualized the moment on a globe whenever two people pressed play on the same song within a millisecond. The song and approx locations of the persons would be shown on a globe, showing like a "connection".

I thought the idea and site was really cool, but at the same time kind of a letdown since you would just see some random people on the globe and there wasn't really any actual connection. From that I got the idea to build a site like Tunemeet, where you could actually connect with people who are listening to the same song right now.

Looking back now in 2025, while the app itself worked and works good, the odds of getting a match are and were quite low as there are are only a handful of users online at the same time. Most of the matches were created in 2020 with users specifically coordinating in Reddit to listen to similar songs, and that was definitely fun while it lasted. I'm keeping the site around as there is still some SEO traffic, and to maybe one day get inspiration to do some marketing and improve it further.

Tech

There are couple of more interesting parts. The first one being the matching engine / logic, that heavily utilizes Redis. On a high-level, when the user starts the matching, we start periodically polling the spotify API to get the track the user is currently listening. That current track for each user is then also stored in Redis. Then to look for matches, we just search if there are matches with that track id in Redis, and if so we pull that data, and match the user.

The second one being the chat, which for some reason always feels kinda interesting, to me at least. The chat and chat presence utilize websockets for a real-time experience, and the messages themselves are stored in Postgres.

Impact / Stats

• 40,000+ users
• 1,100+ matches
• 20,000+ messages sent